Training more but getting slower: a runner’s Ironman, and the 36 minutes it cost him

6 Aug

Reddit post here - https://www.reddit.com/r/IronmanTriathlon/comments/1v6qi9n/im_lake_placid_2026_race_report_and_a_deep_dive/

This week a self-coached athlete put his entire Ironman build on the internet and asked what went wrong. He'd more than doubled his training load between 2025 and 2026, raced the same course both years, and finished 36 minutes slower. Fair question, and he'd done proper work on the data before asking it.

Short answer: he's a runner, and he spent his extra year of training running more.

If you're training more and getting slower, that's usually where I'd start looking. Not the total. The shape of it.

The short version

More than doubled his training load, 11,319 against 4,836. Almost none of the extra went to the bike.
Kept his run schedule as the foundation. Running was the one thing he never cut, and it's the most expensive hour in a triathlon week.
1 to 2 swims a week, slotted into cross-training days. His pool pace improved and his race split still slid from 1:20 to 1:23.
3 standalone rides in a 16 week build. Everything else was a weekend brick, so no ride went long enough to build back-half durability.
Raced a marathon 7.9 weeks out, at the same heart rate as his April marathon. It swallowed the last real bike block.
Biggest run week of the build, 3 weeks out. 70 miles against a normal ceiling near 52, so the taper spent itself absorbing that.

Here's the bit that frames everything. Of the 36 minutes he lost, about 26 came out of the bike split. The run gave up another 9.

So the guy whose best discipline is running lost time on the run. That's not bad luck.

He's a runner. That's the whole story

Ten years of running behind him. A marathon engine. He came into triathlon with the leg most age-groupers spend years trying to build.

And then he trained as though that leg was the thing that needed work.

For a runner, the Ironman marathon is a freshness problem. You've already got the fitness. The question your race asks is whether you arrive at T2 with legs left to use the engine you brought.

He didn't. He described feeling flat on the bike, getting slower and slower over 112 miles, and starting the run already defeated. Then he ran 3:34 off a 3:25 the year before.

That's the whole thing in one number. His weapon was still there. He just handed it a race it couldn't win.

So if you're a runner doing long course, the job is to protect the run, and you protect it almost everywhere except in the run.

Why more training made him slower

Two reasons, and they stack.

First, a training load score is one number, and one number can't tell you where the work went. His went from 4,836 to 11,319. His standalone bike rides went from 5 to 3. Both of those are true at the same time, and only one of them decides an Ironman.

There's a second-order version of this worth knowing, because it's the reason load numbers drift. TSS scales with the square of intensity, so intensity buys load far faster per hour than duration does. An hour at threshold is 100 TSS. That same hour ridden easy at 0.65 is about 42. Double the intensity factor and you quadruple the score for the same hour.

Duration still wins on the total, to be clear about it. 3 hours at 0.65 comes out around 127 TSS and beats the hour at threshold comfortably. The catch is what happens when your hours are capped. With 90 minutes to spend, the quickest way to make the chart climb is to ride harder, and a plan steered by a load number quietly drifts that way across a block.

He wasn't chasing a number. The effect on his week was the same though. His chart went up and to the right while his time under load on the bike stood still.

Second, the extra load landed in the discipline with the highest price tag.

More running is the most expensive hour in triathlon

Every running hour is paid for in tissue. Landing forces, eccentric loading, micro-damage, and a recovery bill that comes due 24 to 48 hours later, usually in your legs on the bike.

Swimming costs you almost nothing mechanically. Cycling sits in between.

So when a runner adds volume by running, he's buying aerobic fitness he already has, at the highest price on the menu, and paying for it with the freshness his bike sessions needed.

I'm not anti-running. I coach runners. But for an athlete with a decade of run base going long course, extra run volume is close to the worst-value hour in the week. Hold the run. Protect the long run. And put the growth somewhere else.

Swim more than feels reasonable

This is the one I'd change first, and it's the one most self-coached triathletes push back on.

He had 1 to 2 swims a week, slotted into days he'd already set aside for cross-training. His pool numbers improved, which tells you the sessions were decent. His race swim still went backwards, 1:20 to 1:23, because 1 to 2 swims a week doesn't hold a swim together over 16 weeks.

Here's what I've watched happen across my roster, and I'll be honest that it's a practitioner read rather than something I can hand you a paper for. Adding swim volume seems to carry across the whole day better than adding run volume does. Athletes who go from 2 swims to 4 tend to come back fitter everywhere, and the ones who go from 4 runs to 6 tend to come back tired everywhere.

Some of that is obvious. It's aerobic work with no landing forces, so you can stack it without a recovery bill. Some of it is the upper body and postural endurance that keeps you holding an aero position at hour 4 instead of sitting up. And some of it is just that swim frequency is the only thing that moves a swim, so the returns are steep when you're only in twice.

If I'd had him last year, swim goes to 3 or 4 sessions a week and it comes out of run volume, not on top of it.

Bricks: fewer, and shorter

He did 5 standalone rides in his 2025 build and 3 in 2026. Everything else was a weekend brick.

I get why. Demanding job, and a brick gets bike and run done in one slot. Most of my athletes have made some version of that trade.

The trouble is a brick caps your ride at whatever leaves room for the run. Usually 2 to 3 hours. In race-specific phase I want an Ironman athlete doing 4 to 5 hour rides at 75 to 82 percent of FTP with race fuelling rehearsed on every one. He never once got a session like that.

His race day describes it perfectly: slowing all the way across the bike leg at a lower heart rate than the year he raced well. That's the tell. Go out too hard when you're fresh and your heart rate runs high. Ride into a hole your training never dug you out of, and heart rate sits low while the watts just aren't available.

Bricks are worth doing. Keep one a week, keep it short, 20 to 30 minutes off the bike. Their job is teaching your legs the transition, and 30 minutes teaches that as well as 90 does. What builds an Ironman bike leg is the long standalone ride, and there's nothing that substitutes for hour 4. I've written more on that in durability in endurance sports.

The marathon 7.9 weeks out

He raced Cape Town at close to full effort, 7.9 weeks before Lake Placid. In 2025 the equivalent slot was an easy aerobic long run, nearly 12 weeks clear of his last hard marathon.

Budget 2 to 4 weeks after a race-effort marathon before real quality comes back. Where an athlete lands in that range is a coping question, not a birthday question. I've got athletes in their 50s who bounce faster than athletes 20 years younger, so I set it off what they've actually shown me.

Call it 3 for him. That leaves about 4 usable weeks of race-specific bike work in a phase that should run 8 to 10. He didn't skip his bike block. It got eaten by a marathon.

Race the marathon if you love racing marathons. Put it 12 or more weeks out, where it sits in base and the recovery is free.

Your biggest run week doesn't belong 3 weeks out

The other thing sitting in his data is a 70 mile run week, 3 weeks before race day, against a normal ceiling around 52.

Taper has one job: shed accumulated fatigue while holding intensity, cutting volume 40 to 60 percent over the final 2 to 3 weeks. It can't do that and absorb the biggest week of your whole build at the same time.

Peak volume wants to sit 5 to 6 weeks out at the latest. And for a runner, I'd question whether that 70 mile week needed to exist at all.

What TSB actually tells you on race morning

He ran a fitness and fatigue model over the data and found his form score on race morning was +25, up from +6 the year he raced well. He read that as evidence he should have been fresher, so fatigue couldn't be the culprit.

Two things.

I want TSB between +5 and +20 on Ironman race morning. Above that, athletes routinely turn up flat, with legs that have forgotten what work feels like. His good race sat at +6. His bad one sat outside the band. The number was pointing straight at the problem.

And hold it loosely regardless. A form score built off Strava Relative Effort isn't TSS calculated from tested threshold pace and power. More to the point, the model can't tell run fitness from bike fitness. A +25 sitting on a run-heavy base says nothing useful about how ready your legs were for 6 hours in the aero bars.

Same goes for a fresh FTP test, by the way. Brilliant anchor for setting zones and tracking a build. Poor oracle for what you'll hold at hour 5.

The thing I'd actually want to see

If he were mine, the first thing I'd ask for is his time in zones.

He's given us total load, discipline volumes, and a form score. None of those tell me how he trained, only how much. I'd want the split: what percentage of his hours sat in easy aerobic, what sat in the moderate middle, what was genuinely hard.

My hunch, and it's only a hunch, is that a run-led build with everything else squeezed around it ends up with far too much time in the middle. Not easy enough to be free, not hard enough to build a ceiling. It's the most common shape I see in self-coached athletes, and it's exactly the shape a load score can't show you.

Two minutes in TrainingPeaks would answer it.

The test that would have caught all of this in June

The frustrating part is that none of this needed to be a surprise on race day.

The durability test I run with Ironman athletes goes inside a single 4 to 5 hour ride. Two matched blocks at race power, around 0.8 IF, 10 to 12 minutes each, same cadence and position. First one at the 50 to 60 minute mark, second in the final hour, around hour 4.

At 0.8 IF the athlete holds the watts both times, so what you're reading is the internal cost: how far heart rate climbs and RPE rises from the early block to the late one. Under 5 percent and your back half is there. Over 10 percent and you back off and rebuild, because race day will find it.

Both blocks sit inside one ride on purpose. Sleep, heat and fuelling stay constant, so a 3 to 5 percent signal doesn't drown in noise.

Run that at week 6 and you get to fix it. He found out at mile 80.

The 5 things I'd change

Swim goes to 3 or 4 sessions a week. Cheapest aerobic hours he owns, and the only way a swim split moves.

Two standalone bike days, one of them long. 4.5 to 5 hours steady through the final 10 weeks, race fuelling rehearsed every time.

Bricks stay, but short. One a week, 20 to 30 minutes off the bike. Enough to teach the legs, not enough to eat the ride.

Run volume held, not grown. Protect the long run, keep the frequency, and stop feeding hours into the discipline he's already good at. Peak week no later than 5 to 6 weeks out.

Last hard marathon 12 weeks out or more.

He's got a serious run engine sitting on an underbuilt bike and a part-time swim. And 11:56 self-coached around a demanding job is properly good going. The 36 minutes are still there waiting for him, and I'd bet a good chunk more on top.

Common questions

Why am I training more but getting slower?

Usually because the extra load went where the race doesn't reward it. A training load score is one number and it can't tell you where the hours went, so it can double while the sessions that decide your race stand still. Check whether your long ride actually got longer.

I'm a runner doing an Ironman. Should I run more?

Probably not. If you've already got a run base, extra run volume buys fitness you have at the highest recovery cost available. Hold your run, protect the long run, and put the growth into swimming and long rides so you reach T2 with legs.

Does swimming more help your bike and run?

In my experience, yes, more reliably than adding run volume does. It's aerobic work with no landing forces, so it stacks without a recovery bill, and the upper-body endurance holds your aero position together late in a long ride.

How many long rides do I need before an Ironman?

Aim for a weekly long ride reaching 4 to 5 hours through the final 10 weeks, at 75 to 82 percent of FTP with race nutrition practised. Fewer than 6 of those in a build and the back half of your bike leg is guesswork.

Do brick workouts replace long rides?

No. A brick caps your ride at whatever leaves room for the run, usually 2 to 3 hours. Keep one short brick a week, 20 to 30 minutes off the bike. Ironman bike durability comes from long standalone rides.

Can I race a marathon during an Ironman build?

Yes, if it sits 12 or more weeks out. A race-effort marathon costs 2 to 4 weeks before real quality returns, so anything inside 10 weeks eats the race-specific bike block that decides your Ironman.

What should my TSB be on Ironman race morning?

I target +5 to +20. Higher and athletes often start flat. Treat it as one input, because a form score is only as good as the threshold values and load data feeding it.

About Mark

I'm Mark Bowstead, a triathlon and Ironman coach based in Auckland, and a former professional triathlete (2011 to 2020). I coach a small roster of time-pressured age-groupers 1-on-1 through Ironman and 70.3 builds, online via TrainingPeaks, with athletes across New Zealand and Australia. Every block gets built around real hours, real recovery, and a real job.

If you're putting the work in and the results have stalled, get in touch and I'll tell you the one thing I'd change.

Mark Bowstead